Summary: use shorter instruction sequences for atomic add and atomic exchange when possible.
Reviewed-by: kvn, jrose

Summary: Fixed vectors alignment when several arrays are accessed in one loop.
Reviewed-by: roland, twisti

Summary: Save whole XMM/YMM registers in safepoint interrupt handler.
Reviewed-by: roland, twisti

Reviewed-by: kvn, jrose, bdelsart

Summary: missing test for T_METADATA leads to incorrect register allocation.
Reviewed-by: kvn

Summary: There was a rogue os::breakpoint() call in log_dependency left over from the jsr292 merge.  Also changed verify_oop() calls for metadata to verify_{method,klass}_ptr.
Reviewed-by: kvn, twisti

Summary: C1 needs knowledge of T_METADATA at the LIR level.
Reviewed-by: kvn, coleenp

Summary: Remove PermGen, allocate meta-data in metaspace linked to class loaders, rewrite GC walking, rewrite and rename metadata to be C++ classes
Reviewed-by: jmasa, stefank, never, coleenp, kvn, brutisso, mgerdin, dholmes, jrose, twisti, roland
Contributed-by: Njmasa &lt;jon.masamitsu@oracle.com&gt;, stefank &lt;stefan.karlsson@oracle.com&gt;, mgerdin &lt;mikael.gerdin@oracle.com&gt;, never <tom.rodriguez@oracle.com>

Reviewed-by: kvn, dholmes, coleenp
Contributed-by: NTao Mao <tao.mao@oracle.com>

Summary: G1 does not appear to be immune to the issue described in CR 6948537 and increasing the size of old-generation PLABs appears to increase the liklihood of seeing the issue. Extend the fix for 6948537 to G1's BlockOffsetTable.
Reviewed-by: brutisso, jmasa

Summary: In C2 add software membar after load from Reference.referent field to prevent commoning of loads across safepoint since GC can change its value. In C1 always generate Reference.get() intrinsic.
Reviewed-by: roland, twisti, dholmes, johnc

Summary: Added asm encoding and mach nodes for vector arithmetic instructions on x86.
Reviewed-by: roland

Summary: Check kstat data for SPARC-M.
Reviewed-by: roland

6984705: JSR 292 method handle creation should not go through JNI
Summary: remove assembly code for JDK 7 chained method handles
Reviewed-by: jrose, twisti, kvn, mhaupt
Contributed-by: NJohn Rose &lt;john.r.rose@oracle.com&gt;, Christian Thalinger &lt;christian.thalinger@oracle.com&gt;, Michael Haupt <michael.haupt@oracle.com>

Summary: renamed mach nodes which use scalar AVX instructions, added integer vectors shuffling instructions
Reviewed-by: twisti

Summary: disable vectorization of a memory access with more elements per vector than one which is used for alignment on sparc
Reviewed-by: twisti

Summary: Increase vector size up to 256-bits for YMM AVX registers on x86.
Reviewed-by: never, twisti, roland

Reviewed-by: kvn

Reviewed-by: kvn, twisti
Contributed-by: NKrystal Mok <sajia@taobao.com>

Summary: increase precision on x86 for the steps of the computation of exp and pow.
Reviewed-by: kvn

Reviewed-by: kvn

Summary: Eliminate constMethod's _method backpointer to methodOop, and move the _constant field from methodOop to constMethod.
Reviewed-by: roland, bdelsart, kamg

Summary: bad test for parity of y in pow(x,y) (c1, interpreter)
Reviewed-by: kvn, twisti

Summary: use x87 instructions to implement exp() and pow() in interpreter/c1/c2.
Reviewed-by: kvn, never, twisti

Summary: Allow a version of String class that doesn't have count and offset fields.
Reviewed-by: never, coleenp

Reviewed-by: chrisphi, dholmes
Contributed-by: NAndrew Dinn <adinn@redhat.com>

Summary: x64 C1 needs to zero upper 32bits when doing l2i conversion
Reviewed-by: never, kvn

Reviewed-by: kvn, iveresov

Reviewed-by: kvn, never
Contributed-by: nils.eliasson@oracle.com

Summary: Object pointer is pushed more than once on stack, where GC doesn't expect it.
Reviewed-by: coleenp, kvn
Contributed-by: axel.siebenborn@sap.com

Summary: Breakpoint needs to do 64-bit compare for pointers on sparcv9
Reviewed-by: coleenp, never
Contributed-by: dean.long@oracle.com

Reviewed-by: kvn, twisti

Reviewed-by: kvn

7148486: At a method handle call returning with an exception may call the runtime with misaligned stack (x64)
Summary: stack must be realigned when calling the runtime for exception propagation at a call.
Reviewed-by: kvn, never

Summary: check stack alignment on VM entry on x64.
Reviewed-by: kvn, never

Summary: Issue storestore barrier before constructor return if the constructor write final field.
Reviewed-by: dholmes, jrose, roland, coleenp
Contributed-by: NJiangli Zhou <jiangli.zhou@oracle.com>

7147064: assert(allocates2(pc)) failed: not in CodeBuffer memory: 0xffffffff778d9d60 <= 0xffffffff778da69c
Summary: Increase size of deopt_blob and uncommon_trap_blob by size of stack bang code (SPARC).
Reviewed-by: azeemj, iveresov, never, phh

7147064: assert(allocates2(pc)) failed: not in CodeBuffer memory: 0xffffffff778d9d60 <= 0xffffffff778da69c
Summary: Increase size of deopt_blob and uncommon_trap_blob by size of stack bang code (SPARC).
Reviewed-by: azeemj, iveresov, never, phh

Summary: Change OSThread to use mach thread_t
Reviewed-by: phh, dcubed

Summary: Replace call_epilog() encoding with macroassembler use. Moved duplicated code to x86.ad. Fixed return_addr() definition.
Reviewed-by: never